Paper Modelling of Ria Formosa v 7

(1)

Biogeochemical Modelling of Ria Formosa (South Portugal)

Duarte, P.

1

*, Azevedo, B.

1

, Guerreiro, M.

1

, Ribeiro, C.

1

, Bandeira, R., Pereira, A.

1

, Falcão,

M.

2

, Serpa, D

2

& Reia, J.

3

1

Center for Modelling and Analysis of Environmental Systems, Faculty of Science and

Technology, Fernando Pessoa University, Praça 9 de Abril 349, P-4200 Porto, Portugal.

pduarte@ufp.pt

2

Portuguese Marine Research Institute (IPIMAR), CRIP Sul, Av. 5 de Outubro, 2700 Olhão,

Portugal

3

Parque Natural da Ria Formosa, Centro de Educação Ambiental do Marim, Quelfes, 8700

Olhão, Portugal

*Author for correspondence

Abstract

Ria Formosa is a large (c.a. 100 km2_{) mesotidal lagunary system with intertidal areas with conflicting uses}

(2)

1. Introduction

The European Water Framework Directive (WFD) (EU, 2000) introduced important changes on the way

water is managed in European Union countries. One of the most important aspects of this directive is the

recognition of the close link between watersheds and coastal waters, namely by defining “River basin

district” – made up of river networks, groundwater and associated coastal waters - as the main river basin

management unit (EU, 2000). This fact is in line with an increasing tendency to link watershed,

hydrodynamic and water quality models (e.g. Park et al., 2003; Plus et al., in 2006 ). The development and

implementations of such integrated approaches is one of the main goals of the European Union DITTY

project (Development of an information technology tool for the management of Southern European lagoons

under the influence of river-basin runoff), where watershed and coastal lagoon models are being applied to

five different southern European ecosystems: Ria Formosa (Portugal), Mar Menor (Spain), Thau lagoon

(France), Sacca Di Goro (Italy) and Gulf of Gera (Greece). Several technical reports are available at the

project web site (http://www.dittyproject.org/). The general approach is to use an offline coupling of

watershed and lagoon models, with the former producing forcing conditions for the latter in terms of river

flows, nutrient and suspended matter loads (Plus et al., in 2006 ).

One of the consequences of the WFD is the need for involved countries to produce studies synthesising the

state of their waters and applying the classification scheme defined in the Directive. Within this classification

scheme, coastal lagoons classify as “Surface water” and typically as “Transitional water” or “Coastal water”,

depending on whether they are substantially influenced by freshwater flows or not.

At present, there is no general agreement about which models to use to simulate watersheds and coastal

(3)

several authors, e.g. Chapelle et al. (2005a). Over the last years, there has been an increasing tendency to

couple hydrodynamic and biogeochemical models in a clear recognition of the importance of incorporating

the feedbacks between physical, chemical and biological processes.

Ria Formosa is a natural park and one of the largest Portuguese coastal lagoons, where many conflicting uses

coexist such as fisheries, aquaculture, harbour activities, tourism and nature conservation. The watershed

draining to this coastal lagoon flows mostly though agricultural lands, where there has been some intensive

use of fertilizers. Management of this coastal ecosystem involves several institutions such as the Natural

Park Authority, several municipalities and the Portuguese Navy. Within the scope of the above mentioned

DITTY project, several possible management scenarios were defined by the Natural Park Authority, that are

being evaluated from the environmental and economic point of view, by using an hydrologic model for the

watershed and a coupled hydrodynamic-ecological model for the lagoon. This work represents the first

approach to this scenario analysis and its objectives are:

(i) Analyse management scenarios related to changes in lagoon bathymetry and their potential

effects on system dynamics.

(ii) Evaluate the relative importance of land drainage, WTP plants and water exchanges, across the

lagoon inlets, for nutrient and suspended matter dynamics;

This study is not a complete assessment of the consequences of the scenarios referred above or of nutrient

and suspended matter discharges, but solely a first attempt to approach their effects at the lagoon scale.

2. Methodology

2.1 Site description

(4)

Ria Formosa watershed is located at the Southernmost part of Portugal (Fig. 1). The origin of its rivers is

mostly in the Caldeirão mountain range and its water courses drain perpendicular to the South in the

direction of the Atlantic Ocean. Most of the rivers are ephemeral with no runoff or very reduced runoff

during part of the year, between June and December. The Ria Formosa basin has an area of 864.26 km2_{, a}

perimeter of 165.99 km, with a maximum altitude of 522 m, draining to the ocean, and an average altitude of

112m, with an average slope of 17% (MAOT, 2000).

Based on annual and monthly data there seems to be an increase in irregularity in annual precipitation in the

basin, being the average annual precipitation value between 600 and 800 mm. The most wet month of the

year is December with about 17% of total annual precipitation, followed by November and January (about

15%). The driest months are July and August with less than 1% of annual precipitation. As far as maximum

daily annual precipitation, for a return period of 2 years, the value is approximately 55 mm, whereas for a

100 years return period it is 132 mm.

2.1.2 Coastal lagoon

Ria Formosa is a shallow mesotidal, eurihaline lagoon located at the south of Portugal (Algarve coast) with a

wet area of 105 km2 (Fig. 1), classified as “Coastal waters” (INAG, 2005) within the scope of the Water

Framework Directive (EU, 2000). The lagoon has several channels and a large intertidal area which

corresponds roughly to 50% of the total area, mostly covered by sand, muddy sand-flats and salt marshes.

The intertidal area is exposed to the atmosphere for several hours, over each semi-diurnal tidal period, due to

its gentle slopes. Tidal amplitude varies from 1 to 3.5 meters and the mean water depth is 3.5 m (Falcão et

al., 2003).

(5)

In this work the SWAT model was used to calculate river flows to force an ecological model of the lagoon.

SWAT, acronym for Soil Water Assessment Tool is a model developed by the USDA Agricultural Research

Service to predict the impact of land management practices on water, sediment, and agricultural chemical

yields in large complex watersheds with varying soils, land use and management conditions over long

periods of time. It is a continuous time model, not designed to simulate detailed, single-event flood routing

(Neitsch et al., 2002).

Rainfall data used in this project is freely available from INAG, which is the Portuguese Water Institute

(www.inag.pt). There are five rain gauges within the Ria Formosa basin area with daily, monthly, yearly and

maximum 24 hour precipitation records. Daily rainfall records were used to run the SWAT model. This

model allows for missing records and uses a weather generator to fill in for these gaps.

Most water that goes into the soil is used by plants through transpiration. Nonetheless, water can percolate

through the soil until it reaches the aquifer and recharge it. Water may even move laterally in the profile and

contribute to stream flow. Therefore, an accurate representation of soil characteristics is important for a

reliable output of the SWAT model. Soils data were obtained from Atlas do Ambiente (IA, 2005) as shape

files to be used by ArcGis. Associated soil characteristics were obtained from a 1965 publication (Cardoso,

1965) and were inserted in the SWAT data base. The soil types present in the Ria Formosa basin are

Cambisols, Fluvisols, Lithosols, Luvisols, and Solonchaks, being Lithosols predominant in the upper basin

and Cambisols and Luvisols in the lower region. As far as the coastal system itself, the predominant soil type

is Solonchak (PROCESL et al., 2000).

Ria Formosa basin has a wide variety of land use classes. Land use data at a 1:25 000 scale was obtained

from the Corine Land Cover maps. There are about 100 different land use classes being divided among six

major groups: 1) urban, 2) agricultural, 3) forest, 4) rangeland and pastures, 5) wetlands and 6) water bodies.

(6)

were further divided for better description in the Ria Formosa basin and the result was a total of eleven land

use classes in the basin (Fig. 2).

As indicated by Neitsch (2002), calibration of a SWAT model run can be divided into several steps:

• water balance and stream flow

• sediment

• nutrients

For the purposes of this work, calibration was performed on stream flow only. Sediment and nutrient loads,

for forcing the lagoon model, were computed from flows and measured concentrations. One stream flow

gauge was used for calibration, Bodega, being the one with most data records. Coiro da Burra has less than

one year of monthly flows, including missing data in the data set, and therefore it was neglected in the

calibration. Curral Boieiros was used for the validation of SWAT parameters (Fig. 1).

As suggested by Neitsch (2002), “calibration for water balance and stream flow should be first done for

average annual conditions, and once the model is calibrated for average annual conditions, the user can shift

to monthly or daily records to fine-tune the calibration” and, therefore, calibration was performed in this

order.

The model was calibrated for annual volume using the data set from the Bodega streamflow gauge, in order

to have some understanding of the actual conditions in the watershed. Calibration was performed manually,

by slightly changing land use and soil variables.

The output from SWAT annual runs is in civil years, rather than water years and that was the time span used

for the analysis. Annual stream flow data published by INAG (www.inag.pt) is in water years, and therefore,

for the annual analysis, monthly data was used for calculation of annual flows, simply adding up all monthly

flows within a civil year. Bodega data set has no missing records from 1953 to 1982, on 1984, and from 1986

(7)

In order to calibrate stream flow, the curve number parameter was adjusted until modelled surface flow

values were approximately the same as stream flow records. This value was adjusted within reasonable limits

for watershed soil, land use and management characteristics. Further adjustment of available soil water

capacity was needed and was performed also within reasonable limits. Most streams are ephemeral,

including the ones in which the stream flow gauges are inserted, being baseflow calibration difficult to

perform.

The overall correspondence between data records and modelled values was analysed using Model II linear

regression analysis, as suggested by Laws and Archie (1981), with the major axis regression method as

recommended by Mesplé et al. (1996) and described in Sokal and Rohlf (1995). ANOVA was used to test

significance of slopes and y-intercepts, as well as the variance explained by the model.

2.3 Hydrodynamic and biogeochemical modelling of the lagunary system

The ecological model implemented in this work is a two dimensional vertically integrated model based on a

finite difference staggered grid (100 m resolution in the present case), coupling hydrodynamic,

thermodynamic and biogeochemical processes. It calculates the velocity field with the equations of motion

and the equation of continuity (Knauss, 1997) and solves the transport equation for all water columns

variables:

( ) ( )

2 2

uS vS

dS S S

Sources Sinks A_x A_y

dt x y _x _y

∂ ∂ _∂ _∂

+ + = + + −

∂ ∂ _∂ _∂ (1)

(8)

u and v - current speeds in x (West-East) and y (South-North) directions (m s-1); A – Coefficient of eddy

diffusivity (m2_s-1_);

S – A conservative (Sources and Sinks are null) or a non conservative variable in the

respective concentration units.

Calculated biogeochemical processes provide the values for the Sources and Sinks terms of equation 1 at

each grid cell.

In the present model, water circulation is forced by tidal height variability and river discharges at sea and

river boundaries, respectively. Tidal height is calculated from the harmonics of the Faro-Olhão harbour

reported in SHOM (1984). The 2D solution for the Navier-Stocks equations is the same described in Neves

(1985), using an ADI (alternating direction implicit) scheme (Dyke, 2001). One important feature of this

model is to include a wet-drying scheme to avoid numerical errors when intertidal areas run out of water. As

described by the previous author, this consists in interrupting flows in those grids cells where water level

drops below a critical value (5 cm in the present case). To guarantee that these cells may be refilled again,

they are considered in the calculations when one of the neighbour cells has a higher water level, allowing for

water to be driven into the “dry” cell by the pressure gradient force.

Water temperature is calculated from standard formulations described in Brock (1981) and Portela & Neves

(1994). Water column biogeochemistry is simulated according to Chapelle (1995) for nitrogen, phosphorus

and oxygen. Processes such as mineralization of organic matter, nitrification and denitrification were

considered for nitrogen. Total and organic particulate matter concentrations (TPM and POM, respectively)

are simulated following Duarte et al. (2003). Particulate organic matter (POM) is mineralized to ammonium

nitrogen as described in the previous author. Oxygen is consumed in mineralization and nitrification and

exchanged across the air-water interface. For more details on the ecological model and a complete listing of

equations and parameters refer to Duarte et al. (2005) and Chapelle et al. (2005a and b). For macroalgae, the

work of Serpa (2004) was used and for the sea grass Zostera noltii, the work of Martin et al. (2003) was

(9)

The model was implemented with EcoDynamo (Pereira & Duarte, 2005) – an object oriented software

written in C++. Table 1 summarizes the objects implemented and their corresponding variables and

processes. Each object simulates several variables and processes and corresponds roughly to the usual

understanding of a sub-model. However, objects have several specific properties that make them very

suitable for modelling, such as modularity, inheritance and polymorphism (Ferreira, 1995).

In EcoDynamo, available objects may be plugged in and out though the model interface, to evaluate the

relative importance of different variables and processes on model solutions. There are two main different

running modes – one with an online coupling of hydrodynamic and biogeochemical processes and another

with an offline coupling. The latter uses previously obtained and time integrated (for 5 minute periods) data

series of current velocities with the hydrodynamic object, to transport water properties among model grid

cells. This allows for a faster simulation, avoiding the computation overhead of hydrodynamic processes and

the small time steps generally required. This simplified mode was used in the present work. Whereas “online

coupling” needs a 3 s time step for stability restrictions, mostly because of very low depths over intertidal

areas, the offline simulations may use a time step of up to 30 s. In fact, a variable time step is used, so that

sites where instabilities may arise are resolved with more detail and properly time integrated with neighbour

cells. Instabilities generally occur when the volume in a cell is very low. In this case, if the time step is not

small enough, the computed flow across one of the cell “walls” times the time step, may be larger than cell

volume. When calculating transport of salt or any other property, this situation may lead to the violation of

mass conservation. The algorithm consists in resolving with more detail these “critical cells” and their

interactions with neighbour cells, finding a time step small enough to prevent mass conservation violations.

In the current model there are no feedbacks from biogeochemistry towards hydrodynamic processes. This is

generally true in barotropic models. In baroclinic simulations, water temperature and density may be

influenced by water turbidity that changes, among other things, as a result of phytoplankton concentration

(10)

The simulations analysed in the present work were not carried out with full model complexity. Only the

“Wind”, “Air temperature”, “Water temperature”, “Tide”, Hydrodynamic 2D”, “Dissolved substances” and

“Suspended matter” objects were considered (cf. – Table 1). The simulations were designed to understand

the relative contribution of specific processes within the western part of Ria Formosa (Fig. 1), according to

the objectives referred above (cf. – Introduction), and may be viewed as a “virtual” experiment, with a few

simplifying assumptions. Their accuracy depends mostly on the quality of the hydrodynamic simulation.

Therefore, the calibration and validation of the hydrodynamic object is analysed in this study and based on

current velocity data collected by the Portuguese Hydrographic Institute in 2001 (IH, 2001) at a number of

stations (Fig. 1), over periods of several days (not necessarily coincident among different stations), between

January and March 2001. Water quality data for the rivers draining to Ria Formosa, inside the lagoon system

and at the sea boundaries, were obtained in several works carried out by the Marine Research Institute

(Falcão & Vale, 1990; Vale et al., 1992; Falcão & Vale, 1995; Falcão, 1997, Falcão & Vale, 1998; MAOT,

2000; Falcão & Vale, 2003; Newton et al., 2004). A study using the full model complexity, including its

calibration and validation is currently being carried out and it will be the subject of an upcoming paper.

A first set of hydrodynamic simulations was carried out for the same period used in model

calibration/validation to analyse the effects of several scenarios related to changes in lagoon bathymetry and

depicted in Fig 3. These scenarios were defined by Ria Formosa Natural Park staff on the basis of past

dredging activities and anticipating the need to improve navigation conditions within some of the main

channels. Results from the various simulations were analysed by comparing obtained water washout times,

time integrated flows across the inlets and current velocities at points used for model calibration/validation.

Water residence times were estimated by “filling” the lagoon with a conservative tracer and running the

model until its “washout” to the sea.

A second set of simulations (Table 2) was performed to understand the relative importance, on Ria Formosa

(11)

conservative and some non-conservative processes. In this case, the offline mode was adopted (see above).

Two river flow regimes were considered – winter and summer - estimated with the SWAT model (see

above). Comparing results obtained with different river discharge regimes (nearly zero discharge for

the summer situation) or/and WWTP discharges permits to understand the relative contribution of land drainage and WWTPs to water column nutrient and suspended matter concentrations. Contrasting conservative with non-conservative simulations allows understanding the relative importance of

water column biogeochemistry in explaining variability of those variables. When “Suspended matter” object is treated as conservative, POM is not mineralized to ammonium and phosphate. When “Dissolved substances” is treated as conservative, ammonium may increase due to POM mineralization, but

nitrification and denitrification do not occur. In all simulations, the model was initialized with values well

within the range of those observed in Ria Formosa and obtained from a data base created within the DITTY

project (http://www.dittyproject.org/). Both simulation sets were run to simulate a period of one month.

3. Results

3.1 Hydrologic modelling

As far as annual flow, SWAT adequately models it, as can be observed from Fig. 4 and from the results of

Model II linear regression analysis performed. The slope of the Model II regression between data records and

modelled values was not significantly different from one and the y-intercept was significantly different from

zero (p < 0.05). The variance explained by the model was significant (p << 0.05). These results imply that

the model explains a significant proportion of the observed variance. However, the model tends to

overestimate measured annual flows.

Analogous to annual flow, monthly flow is adequately modelled by SWAT, as can be observed from Figs. 5

(12)

significant (p << 0.05) in most months, except the summer months of July, August and September. This is

probably due to the absence of rain during those months in which the model cannot predict the stream flow

well. Slope between data records and modelled values was not significantly different from one and

y-intercept was not significantly different from zero (p < 0.05) except on the summer months.

3.2 Hydrodynamic and biogeochemical modelling of the lagunary system

In Figs. 7 – 10, measured and predicted current velocities are shown for each of the monitoring locations

depicted in Fig. 1 at the western part of Ria Formosa. The visual fit between measurements and observations

is generally good. The slope of the Model II regression between measured and observed values (cf. –

Methodology – Hydrologic modelling) was significantly different (s.d.) from one and the y-intercept was s.d.

from zero (p < 0.05) in almost all simulations. The variance explained by the model was significant (p <<

0.05) in all cases. These results imply that the model explains a significant proportion of the observed

variance. However, it tends to underestimate measured velocities. This is an expected result, because model

velocity predictions correspond to spatially integrated values for each grid cell, whereas measurements are

performed in one point in space, within the channels, where current velocities tend to be higher.

Ideally, two independent data sets should have been used – one for calibration and another for validation.

However, since there was only one dataset available and since the model reproduced observed data relatively

well, without any calibration effort, calibration and validation are here considered together. Furthermore,

changing model parameters locally, such as turbulent diffusivity or bottom drag, seeking for a better model

fit to observed data, would hardly be consistent in future simulations, in a system where bottom

configuration and bathymetry changes so rapidly. Therefore, efforts were mostly directed towards a rigorous

bathymetric description and the determination of the accurate position of all inlets at the time when sampling

(13)

Current speeds range from nearly zero till values in excess of 100 cm s-1_{. Velocity peaks occur both at the}

middle of the ebb and the middle of the flood. This is a normal phenomena in inlets - when current switches

from flood to ebb, the water level is near its peak flood value (Militello & Hughes, 2000). General

circulation patterns within Ria Formosa are shown in Fig. 11, during the flood and during the ebb. Maximum

current velocities are observed at the inlets. During the ebb, water remains only in the main channels.

Residual flow suggests the existence of eddies near the inlets and also close to Faro-Harbour (cf. – Fig. 1).

The comparison of ebb and flood tidal periods, predicted by the model, confirms flood dominance (Table 3).

According to model results, the flood period may be larger than the ebb period by nearly two hours in Ancão,

and Fuzeta-Canal. These patterns may be explained by flow divergence (see below).

The integration of flows across the inlets made possible to estimate their average input-output values for a

period of a month. In Fig. 12, a synthesis of obtained results over the whole Ria shows that the Faro-Olhão

inlet is by far the most important, followed by Armona, “new” and Fuzeta inlets. It is also apparent that the

Faro-Olhão has a larger contribution as an inflow pathway, whereas the remaining ones contribute more as

outflow pathways. The small difference between inflow and outflow total values do not imply any violation

of volume conservation, but solely that during the period considered there was a net exchange of volume

between the Ria and the sea. The results obtained suggest that part of the water that enters the Ria though the

Faro-Olhão inlet is distributed west and eastwards (cf. Fig. 11), probably reducing the flood period in other

areas. The results presented in Table 3 suggest that ebb period is larger than the flood period. This may result

from ebb water taking more time to reach the ocean by outflowing only thought nearby inlets, whereas

during the flood, there seems to be some volume redistribution among different inlets. Water residence time

(considering a 90% washout) ranges from less than one day, near the inlets, to more than two weeks, at the

inner areas, with an average value of 11 days.

Table 4 summarizes the results of the first simulation set (cf. – Methodology - Hydrodynamic and

(14)

on the time necessary for the washout of 50, 90 and 99% of lagoon water, time integrated flows across the

inlets and current velocities at chosen points were investigated.

Table 5 summarizes the results of the second simulation set (cf. – Methodology - Hydrodynamic and

biogeochemical modelling of the lagunary system), to investigate the effects of flow discharges from rivers

and from WWTPs, and of conservative and some non-conservative processes on water quality.

4. Discussion

SWAT model was successfully calibrated and validated except for the summer (dry) months, where the

streams have a low flow or no flow at all. The average river flow draining to “Western” Ria Formosa is c.a.

0.4 m3 s-1, a rather low figure considering the size of the lagoon. The model was used for the Ria Formosa

basin system to generate annual, monthly and daily average flows as input to the lagoon model presented in

this study, after its successful calibration and validation.

Obtained results with the lagoon model (first simulation set, cf. 2.3 and 3.2) show that channel deepening

though dredging operations tends to increase water washout time (cf. Table 4), presumably due to the

corresponding increase in lagoon volume, whereas sand accretion at the “Fortaleza Growing Area” has the

opposite effect. There are some exceptions, but these correspond to less than 1% changes in water residence

times. The “Fuseta Channel” scenario (Fig. 3) exhibits the largest outflow reduction across the “New” and

the “Fuzeta” inlets. This may be viewed as a negative impact, since outflow reduction may increase sand

accumulation within the lagoon. These trends suggest that bathymetric changes in one side of the lagoon may

have impacts tens of km away.

(15)

a poor influence of water column biogeochemical processes. The largest effect on nitrate is explained by its

high concentrations in river water (values in excess of 500 mol N L-1_{). These high nitrate loads may}

probably be explained by intensive use of fertilizers at the extensive agricultural areas drained by the river

network (cf. - Fig. 2). However, the model predicts a rapid decline in nitrate concentrations with distance

from river mouths (Fig. 13). It is noteworthy that ammonium concentrations practically double when TPM or

nutrients are treated as non-conservative (Table 5, simulations 4, 5 and 6), as a result of POM mineralization

or denitrification, respectively. This doubling is much larger than the combined effect of river and WWTP

discharges. It is also relevant to see that WWTP discharges seem to contribute more than river discharges for

ammonium concentrations.

The results presented here are not in full accordance with the classification of Ria Formosa as “Coastal

waters” (cf. – Methodology – Site description). The classification as “Transitional waters”, implying a

substantial influence by freshwater flows (EU, 2000), seem to apply when river discharges are relevant,

namely, in winter months and in the case of nitrate.

Subtidal and intertidal areas of the lagoon are extensively covered by benthic macrophytes, such as

macroalgae (Enteromorpha spp. and Ulva spp.), seagrasses (Zostera sp., Cymodocea nodosa and Ruppia

cirrhosa) and Spartina maritima that dominate the low salt marshes (Falcão, 1997). The inter-tidal areas are

mainly covered by Spartina maritima (8 km2), seagrasses (8.2 km2) and macroalgae mats (2.5 km2) (Aníbal,

1998). From these vegetation cover values, annual production estimates and known Redfield ratios for the

various taxonomic groups, nitrogen and phosphorus daily mean uptakes may be obtained. Regarding

macroalgae, such estimates are reported in Serpa (2004). Concerning Spartina maritima and Zostera noltii

(the dominant seagrass), production estimates are reported in Santos et al. (2000), whereas nitrogen and

phosphorus contents were taken from Valiela (1995). A similar approach was followed for phytoplankton,

(16)

Results obtained are summarized in Table 6, together with daily river nitrogen and phosphorus discharges. It

is noteworthy that the values presented are only approximate, since they do not take into account subtidal

biomasses of benthic species, however, they seem to show that the contribution of river nutrient discharges

to primary production, corresponds roughly to macroalgae nitrogen and phosphorus consumption. They also

suggest that primary producers may be ordered by decreasing production rates and nutrient consumptions as

phytoplankton, Zostera noltii, Spartina maritima and macroalgae. This contradicts results obtained by other

authors in shallow coastal lagoons and bays, where macrolgae production dominates over phytoplankton

(Sfriso et al., 1992; Valiela et al., 1992; McGlathery et al., 2001). The lower phytoplankton production has

been attributed to nutrient competition between macroalgae and phytoplankton (Fong et al., 1993;

Thybo-Christensen & Blackburn, 1993; McGlathery et al., 1997) and to water residence times shorter than

phytoplanktondoubling time (Valiela et al., 1997). This contradiction may be tentatively explained by:

(i) Benthic production does not seem to be macroalgae dominated in Ria Formosa, with rooted

macrophytes playing an important role (Table 6). In fact, macroalgae tend to dominate as

lagoons become eutrophic (Harlin, 1995), which is not the case of Ria Formosa.

(ii) Water residence time is longer than phytoplankton doubling time in Ria Formosa (less than 2

days (Duarte et al., 2003)) and it takes approximately 11 days for a 90% water exchange

between the lagoon and the sea (see above).

Obtained results also suggest the importance of other sources of nutrients than the watershed in Ria Formosa

biogeochemistry, such as nitrogen fixation, inputs from the sea and sediment water interactions.

5. Conclusions and future work

From the results presented and discussed, it may be concluded that scenarios related to changes in lagoon

bathymetry tend to increase lagoon water washout time. This result is explained by an increase in lagoon

(17)

those with a largest impact, since they result in the largest effect on washout time. Changes in lagoon

bathymetry may have impacts at a scale of tens of km away, as seen by changes on average inflows and

outflows across inlets located far away from dredged areas (cf. Table 4). This conclusion justifies the usage

of a lagoon scale model, as in the present work.

Watershed and WWTP contribution is mostly through nitrate loads, whereas WWTP contribution is mostly

through ammonium loads. From a management point of view, dredging operations are important to improve

navigability, but may have a negative impact on water quality if water washout time is increased.

Furthermore, this increase may have important implications in lagoon biogeochemistry due to the apparent

dependence of nutrient cycles on lagoon-sea exchanges.

The results obtained in this work and corresponding conclusions help to understand the relative importance

of the watershed at this part of the river basin district (EU, 2000) called “Ribeiras do Algarve”, suggesting

that the high nitrate concentrations at the river network may have a significant effect on lagoon

concentrations, in spite of the relatively low river flows.

Acknowledgements

This work was supported by DITTY project (Development of an information technology tool for the

management of Southern European lagoons under the influence of river-basin runoff)(EESD Project

EVK3-CT-2002-00084). The authors wish to thank the critics from two anonymous reviwers.

5. References

Aníbal, J., 1998. Impacte da macroepifauna sobre as macroalgas Ulvales (Chlorophyta) na Ria Formosa. MSc Thesis. Coimbra University, 73 pp.

(18)

Cardoso, J.V., 1965. Os solos de Portugal, sua classificação, caracterização e génese. Secretaria de Estado da Agricultura - Direcção Geral dos Serviços Agrícolas, Lisboa.

Chapelle, A., 1995. A preliminary model of nutrient cycling in sediments of a Mediterranean lagoon. Ecological Modelling 80: 131-147.

Chapelle, A., C. Bacher, P. Duarte, A. Fiandrino, L. Galbiati, D. Marinov, J. Martinez, A. Norro, A. Pereira, M. Plus, S. Rodriguez, G. Tsirtsis & J.M. Zaldívar, 2005a. Modelling report. Coastal lagoon modelling: An integrated approach. Available at http://www.dittyproject.org/Reports.asp.

Chapelle, A., P. Duarte, A. Fiandrino, L. Galbiati, D. Marinov, J. Martinez, A. Norro, F. Somma, G.. Tsirtsis, J.M. Zaldívar, et al., 2005b. Comparison between different modelling approaches. Available at http://www.dittyproject.org/Reports.asp.

Duarte, P., R. Meneses, A.J.S. Hawkins, M. Zhu, J. Fang & J. Grant, 2003. Mathematical modelling to assess the carrying capacity for multi-species culture within coastal water. Ecological Modelling 168: 109-143.

Duarte, P., B. Azevedo & A. Pereira, 2005.

Hydrodynamic Modelling of Ria Formosa (South Coast

of Portugal) with EcoDynamo

. DITTY report. Available at http://www.dittyproject.org/Reports.asp.

(19)

European Union (EU), 2000. Directive 2000/60/EC of the European Parliament and of the Council establishing a framework for the Community action in the field of water policy. Official Journal of the European Communities, 22/12/2000.

Falcão, M., 1997. Dinâmica de nutrientes na Ria Formosa: efeitos da interacção da laguna com as suas interfaces na reciclagem do azoto, fósforo e sílica. PhD Thesis, Universidade do Algarve, 223 pp.

Falcão, M. & C. Vale, 1990. Study of the Ria Formosa ecosystem: benthic nutrient remineralization and tidal variability of nutriens in the water. Hydrobiologia 207: 137-146.

Falcão, M. & C. Vale, 1995. Tidal flushing of ammonium from intertidal sediments of Ria Formosa, Portugal. Netherlands Journal of Aquatic Ecology, 29 (3-4): 239-244.

Falcão, M. & C. Vale, 1998. Sediment-water exchanges of ammonium and phosphate in intertidal and subtidal areas of a mesotidal coastal laggon (Ria Formosa). Hydrobiologia 373/374: 193-201.

Falcão, M. & C. Vale, 2003. Nutrient dynamics in a coastal lagoon (Ria Formosa, Portugal): The importance of lagoon-sea water exchanges on the biologic productivity. Ciencias Marinas, 29 (3): 425-433.

(20)

Ferreira J.G., 1995. EcoWin - An Object-oriented Ecological Model for Aquatic Ecosystems. Ecological Modelling 79: 21 – 34.

Fong P, R.M., Donohoe, J.B. Zedler, 1993. Competition with macroalgae and benthic cyanobacterial mats limits phytoplankton abundance in experimental microcosms. Marine Ecology Progress Series 100: 97–102.

Harlin, M.M., 1995. Changes in major plant groups following nutrient enrichment. In: McComb, A.J. (eds.), Eutrophic shallow estuaries and lagoons. CRC Press: 173 – 187.

Instituto da Água (INAG), 2005. Relatório síntese sobre a caracterização das regiões hidrográficas prevista na Directiva-Quadro da água. Instituto da Água, Ministério do Ambiente, do Ordenamento do Território e do Desenvolvimento Regional.

Instituto do Ambiente (IA) (Ministério do Ambiente e do Ordenamento do Território), 2005. Atlas do Ambiente. Available at http://www.iambiente.pt/atlas/est/index.jsp.

Instituto Hidrográfico (IH), 2001. Proj. OC4102/01, Maria Formosa, Relatório Técnico Final, Rel. TF. OC 04/2001, Monitorização Ambiental. Instituto Hidrográfico, Divisão de Oceanografia.

Knauss, J.A., 1997. Introduction to physical oceanography. Prentice-Hall.

(21)

MAOT (Ministério do Ambiente e do Ordenamento do Território), 2000. Plano de Bacia Hidrográfica das Ribeiras do Algarve – Caracterização Geral da Bacia Hidrográfica, 1ª Fase – Análise e Diagnóstico da Situação de Referência, Vol. III.

McGlathery, K.J., I.C.Anderson & A.C.Tyler, 2001. Magnitude and variability of benthic and pelagic metabolism in a temperate coastal lagoon. Marine Ecology Progress Series 216: 1-15.

McGlathery KJ, Krause-Jensen D, Rysgaard S, Christensen PB (1997) Patterns of ammonium uptake within dense mats of the filamentous macroalgae

Chaetomorpha linum

. Aquatic Botany 59:

99–115.

Mesplé, F., M. Trousselier, C. Casellas & P. Legendre, 1996. Evaluation of simple statistical criteria to qualify a simulation. Ecological Modelling 88: 9-18.

Militello, A. & S.A. Hughes, 2000. Circulation patterns at tidal inlets with jetties. US Army Corps of Engineers.

Neitsch, S.L., J.P. Arnold, J.R. Kiniry, R. Srinivasan & J.R. Williams, 2002.

Soil and Water

Assessment Tool. User´s Manual.

Grassland, Soil and Water Research Laboratory, Agricultural

Research Service.

Neves, R.J.J., 1985. Étude expérimentale et modélisation mathématique des circulations transitoire et rédiduelle dans l’estuaire du Sado. Ph.D. Thesis, Université de Liège.

(22)

Park, K., H.-S. Jung, H.-S. Kim, & S.-M. Ahn, 2003. Estuarine and coastal water quality modelling: Concept and a case study in Korea. In: Yu and Bermas (eds.), Determining Environmental Carrying capacity of coastal and marine areas: Progress, constraints, and future options, GEF/UNDP/IMO, PEMSEA: 98 – 114.

Pereira, A. & P. Duarte, 2005. EcoDynamo: Ecological Dynamics Model Application. DITTY report. Available at http://www.dittyproject.org/Reports.asp.

Plus, M., A. Chapelle, A. Ménesguen, J.M. Deslous-Paoli & I. Auby, 2003. Modelling seasonal dynamics of biomasses and nitrogen contents in a seagrass meadow (

Zostera noltii

Hornem.):

application to the Thau lagoon (French Mediterranean coast). Ecological Modelling 161: 213-238.

Plus, M., I. Jeunesseb, F. Bouraouic, J.-M. Zaldívar, A. Chapelle, & P. Lazuref. Modelling water discharges and nutrient inputs into a Mediterranean lagoon. Impact on the primary production. Ecological Modelling 193:69-89

Portela, L.I. & R. Neves, 1994. Modelling temperature distribution in the shallow Tejo estuary. In: Tsakiris and Santos (eds), Advances in Water Resources Technology and Management, Balkema, Rotterdam: 457-463.

(23)

Santos, R., M. Sprung, R. Machás, J. Aníbal, N. Dias, L. Mata, V. Vieira, F. Piedade, L. Pérez-Lloréns, I. Hernández, J. Vergara & G. Peralta, 2000. Produção bentónica e fluxos de matéria orgânica na Ria Formosa, Algarve, Portugal. Final report of the INTERREG II Project (contract nº 12/REG/6196

).

Serpa, D., 2004. Macroalgal (

Enteromorpha spp.

and

Ulva spp

.) Primary Productivity in the Ria

Formosa Lagoon. MSc thesis, Universidade Nova de Liaboa.

Service Hydrographique et Océanographique de la Marine (SHOM), 1984. Table des marées des grands ports du Monde. Service Hydrographique et Océanographique de la Marine.

Sfriso, A., B. Pavoni, A. Marcomini & A.A. Orio, 1992. Macroalgae, nutrient cycles, and pollutants in the Lagoon of Venice. Estuaries 15: 517 – 528.

Sokal, R.R. & F.J. Rohlf, 1995. Biometry, The principles and practice of statistics in biological research. W.H. Freeman and Company.

Thybo-Christesen M & T.H. Blackburn, 1993. Internal N-cycling, measured by 15NH4+ dilution, in

Cladophora

sericea

in a shallow Danish bay. Marine Ecology Progress Series 100: 283–286.

Vale, C., J. Hussenot, M. Falcão & J. Martin, 1992. Dawn-dusk chemical variations near the sediment-water interface of a clam-growth bank of Ria Formosa. Boletim INIP, 71-71h.

(24)

Valiela I, K. Foreman, M. LaMontagne, D. Hersh, J. Costa, P. Peckol, B. DeMeo-Andreson, C. D’Avanzo, M. Babione, C.H. Sham, J. Brawley & K. Lajtha, 1992. Couplings of watersheds and coastal waters: sources and consequences of nutrient enrichment in Waquoit Bay, Massachusetts. Estuaries 15: 443–457.

(25)

Table 1 – EcoDynamo objects implemented for Ria Formosa and respective variables (see text).

Object type

Object name

Object outputs

Wind object Wind speed

Air temperature object Air temperature

Water temperature object Radiative fluxes and balance between water and atmosphere and water temperature

Light intensity object Total and photosynthetically active radiation (PAR) at the surface and at any depth Objects providing forcing

functions

Tide object Tidal height

Hydrodynamic 2D object Sea level, current speed and direction

Sediment biogeochemistry object Pore water dissolved inorganic nitrogen (ammonia, nitrate and nitrite), inorganic phosphorus and oxygen, sediment adsorbed inorganic phophorus, organic phosphorus, nitrogen and carbon

Dissolved substances object Dissolved inorganic nitrogen (ammonia, nitrate and nitrite), inorganic phosphorus and oxygen

Suspended matter object Total particulate matter (TPM), particulate organic matter (POM), carbon (POC), nitrogen (PON),phosphorus (POP) and the water light extinction coefficient

Objects providing state variables

Phytoplankton object Phytoplankton biomass, productivity and cell nutrient quotas

Enteromorpha sp. object Macroalgal biomass, productivity and cell nutrient quotas

Ulva sp. object Macroalgal biomass, productivity and cell nutrient quotas

Zostera noltti object Macrophyte biomass and numbers, cell nutrient quotas and demographic fluxes Clams (Ruditapes decussates) object Clam size, biomass, density,

(26)

Table 2 – Synthesis of second set of simulations analysed in the present work. A total of 12 simulations were carried out. For conservative simulations, a zero value was assumed for all biogeochemical rate constants regarding mineralization, nitrification and denitrification. For non-conservative simulations the values reported in Chapelle (1995) were used with oxygen and temperature limitation (cf. – Methodology – Simulations).

Discharges

Simulation

nº

River discharges

WTP discharges

Type

Conservative

Non-conservative

1 Winter Yes

2 Winter No

3 4

Summer Summer

Yes No

Conservative

5 Winter Yes

6 Winter No

7

8 Summer Summer Yes No

Suspended matter object non-conservative Dissolved

substances object conservative

9 Winter Yes

10 Winter No

11

12 Summer Summer Yes No

Suspended matter object

conservative Dissolved

(27)

Table 3 – Predicted average ebb and flood current velocities and periods at the current meter stations depicted in Fig. 1 for he “Western” Ria Formosa (see text).

Ebb Flood

Station Average current velocity (cm s-1)

Period (h) Average current velocity (cm s-1)

Period (h)

Ancão 17.90 7.16 24.57 5.20

Faro-Harbour 50.69 6.10 39.49 6.06

Olhão-Canal de Marim

32.30 6.72 31.07 5.47

(28)

Table 4 – Summary of time for a 50, 90 and 99% washout of lagoon water and inflow and outflow changes in relation to the validation scenario (see text).

Variations (%)

Washout times

Flows

Scenarios

50%

90%

99%

Inlets

Inflows Outflows

"New Inlet" 64.0 -5.3 Faro-Olhão 8.6 19.5 Armona 12.4 10.2 Ramalhete Channel 1.6 10.0 6.0

Fuseta -12.4 20.4 "New Inlet" 64.0 -8.5 Faro-Olhão -1.9 11.6 Armona 2.8 7.5 Faro-Olhão Inlet 24.4 28.7 13.9

Fuseta -18.0 13.1 "New Inlet" 75.8 -3.5 Faro-Olhão 9.7 18.2 Armona 7.3 7.6 Olhão Channel -0.3 0.6 0.1

Fuseta -8.7 17.0 "New Inlet" 72.8 -3.2 Faro-Olhão 8.3 16.4 Armona 12.1 10.5 Fuseta Channel 44.2 75.1 26.4

Fuseta -46.9 -22.6 "New Inlet" 77.5 -5.9 Faro-Olhão 8.1 15.9 Armona 10.7 8.4 Fortaleza Growing

Area -1.5 -1.4 -2.0

(29)

Table 5 – Summary of simulations described in Table 2. All results are in mol L

-1

for nutrients and mg L

-1

for TPM and POM (see text).

Ammonium

Nitrate

Nitrite

Phosphate

TPM

POM

Simulation

Average

Max

Average

Max

Average

Max Average

Max

Average

Max

Average

Max

1

0.50

4.23

4.24 674.69 0.12

3.01 0.43

17.05 6.06

40.00 0.26

5.53

2

0.36

4.23

4.23 674.69 0.12

3.01 0.41

17.05 6.06

40.00 0.25

5.53

3

0.49

4.23

2.76 674.69 0.12

3.01 0.40

17.05 6.04

40.00 0.25

5.53

4

0.36

4.23

2.74 674.69 0.12

3.01 0.38

17.05 6.04

40.00 0.25

5.53

5

0.95

4.23

4.24 674.69 0.12

3.01 0.44

17.05 6.04

40.00 0.23

5.53

6

0.81

4.23

4.23 674.69 0.12

3.01 0.42

17.05 6.03

40.00 0.23

5.53

7

0.97

4.23

2.76 674.69 0.12

3.01 0.41

17.05 6.02

40.00 0.22

5.53

8

0.82

4.23

2.74 674.69 0.12

3.01 0.39

17.05 6.01

40.00 0.22

5.53

9

0.83 13.49 3.88

674.69 0.12

3.01 0.43

17.05 6.06

40.00 0.26

5.53

10

0.75 13.01 3.8

674.69 0.12

3.01 0.41

17.05 6.06

40.00 0.25

5.53

11

0.62 14.12 2.58

674.69 0.12

3.01 0.40

17.05 6.04

40.00 0.25

5.53

(30)

Table 6 – Estimates of nitrogen and phosphorus daily consumptions by main primary producers in Ria Formosa, from production figures and known Redfield ratios, and river discharges (see text).

Nitrogen (kg d

-1

)

Phosphorus (kg d

-1

)

Spartina maritima

289 - 552 19 – 37

Zostera

noltii

473 - 647 31 - 43

Macroalgae 189 27

Phytoplankton 546 76

(31)

Tavira-Cabanas Tavira-Clube Naval

Fuzeta -Canal

Olhão – Canal de Marim Faro-Harbour Ancão Olhão Faro – main channel X X %

C URRA L B OIE I R O BO DE GA

CO IR O DA B UR RA

N

1 0 00 0 0 10 0 0 0 2 00 0 0 M e t e rs Ribeira de S. Lourenço Rio Gilão Ribeira de Almargem Rio Seco Ribeira de

Biogal Ribeira de

Bela Mandil Ribeira de

Marim Ribeira do Troaco

X X

%

CO IR O DA B UR RA

X X

%

CO IR O DA B UR RA

N

Biogal Ribeira de

Easter

n

Ria

Wester

n

Ria

Tavira-Cabanas Tavira-Clube Naval Fuzeta -Canal Olhão – Canal de Marim Faro-Harbour Ancão Olhão Faro – main channel X X %

CO IR O DA B UR RA

N

Biogal Ribeira de

X X

%

CO IR O DA B UR RA

X X

%

CO IR O DA B UR RA

N

Biogal Ribeira de

Wastewater Treatment Plants

Eastern Ria Western Ria Tavira-Cabanas Tavira-Clube Naval Fuzeta -Canal Olhão – Canal de Marim Faro-Harbour Ancão Olhão Faro – main channel X X %

CO IR O DA B UR RA

N

Biogal Ribeira de

X X

%

CO IR O DA B UR RA

X X

%

CO IR O DA B UR RA

N

Biogal Ribeira de

Easter

n

Ria

Wester

n

Ria

Tavira-Cabanas Tavira-Clube Naval Fuzeta -Canal Olhão – Canal de Marim Faro-Harbour Ancão Olhão Faro – main channel X X %

CO IR O DA B UR RA

N

Biogal Ribeira de

X X

%

CO IR O DA B UR RA

X X

%

CO IR O DA B UR RA

N

Biogal Ribeira de

Wastewater Treatment Plants

Eastern Ria Western Ria

(32)

N

L a n d u s e

1 0 0 0 0 0 1 0 0 0 0 2 0 0 0 0 M e te rs

L a n d u s e A G R L F R S D F R S E F R S T O R C D P I N E R N G B U R M L U R M M U T R N W E T L

Fig. 2 - Land use in Ria Formosa basin. AGRL - Agricultural Land-Generic; FRSD - Forest-deciduous; FRSE - Forest-evergreen; FRST - Forest-mixed; ORCD – Orchard; PINE – Pine; RNGB - Range-brush;

(33)

Ramalhete Channel Dredging to 2 m depth

Olhão Channel Dredging to 8 m Depth

Faro-Olhão Inlet

Fuseta Channel Dredging to 2 m depth

Fortaleza Cultivation Area

(34)

Annual Runoff - Bodega

0 100 200 300 400 500 600 700 800

0.0 100.0 200.0 300.0 400.0 500.0 600.0 700.0

Data records - mm

M

o

d

e

le

d

v

a

lu

e

s

m

(35)

!

" #

$ $

!

%

$ $

!

&' (

$

!

! !

!

" #

$ $

!

%

$ $

!

&' (

$

!

! !

(36)

$

!

& )

$ $

!

* ' #

$ $

!

+ #

$

$ !

, - #

!

#

! $

!

& )

$ $

!

* ' #

$ $

!

+ #

$

$ !

, - #

!

#

!

(37)

0 10 20 30 40 50 60

13 14 15 16 17 18 19 20 21

Time (days)

c

m

s

-1

Simulated velocity Observed velocity

Fig. 7 – Predicted and measured velocities at Ancão (cf. – Fig. 1).

0 10 20 30 40 50 60 70 80 90 100 110

1 2 3 4 5 6 7 8 9

Time (days)

c m s -1 0 10 20 30 40 50 60 70 80 90 100 110

1 2 3 4 5 6 7 8 9

Time (days)

c

m

s

-1

(38)

0 10 20 30 40 50 60 70

14 15 16 17 18 19 20 21

Time (days)

c

m

s

-1

Fig. 9 – Predicted and measured velocities at Faro-Harbour (cf. – Fig. 1).

0 10 20 30 40 50 60 70 80 90 100

2 3 4 5 6 7 8 9 10

Time (days)

c m s -1 0 10 20 30 40 50 60 70 80 90 100

2 3 4 5 6 7 8 9 10

Time (days)

c

m

s

-1

(39)

ms - 1

Ebb

Flood

cms- 1

Ebb

Flood

125

166

ms - 1

Ebb

Flood

cms- 1

Ebb

Flood

ms - 1

Ebb

Flood

cms- 1

Ebb

Flood

125

166

(40)

“

new

”

inlet

“

new

”

inlet

Faro

-

Olhão

Faro

-

Olhão

Armona

Fuzeta

226

142

1768

2033

1082

1133

141

157

“new” inlet

Faro - Olhão

Armona

Fuzeta

336

292

1952

2096

1234

1259

144

210

(41)